The overall goal of this project continues to be the design and synthesis of new antifolates for cancer chemotherapy, with particular emphasis on innovative approaches leading to dihydrofolate reductase (DHFR) inhibitors whose mode of action is designed to set them favorably apart from 'classical' inhibitors such as aminopterin (AMT), methotrexate (MTX), and edatrexate (EDX). SPECIFIC AIM 1: To extend and complete the currently ongoing work on tightly DHFR bound and efficiently transported but nonpolyglutamated analogs, of which the lead compound, N-alpha-(4-amino-4deoxypteroyl)- N-delta-hemiphthaloyl-L-ornithine (PT523, NSC 633713), was recently selected by the National Cancer Institute for accelerated preclinical development and Phase I testing as part of its new RAID program. SPECIFIC AIM 2: To synthesize a series of DHFR inhibitors which are good FPOS substrates but are limited to the addition of only one glutamyl residue. This will be accomplished by blocking free rotation of the side chain amide bond via the use of a bridge as has been done in the case of the potent thymidylate synthase inhibitor GW1843. These compounds will represent an intermediate class of DHFR inhibitors between those which are metabolized to longchain polyglutamates (e.g., MTX, EDX) and those which are not polyglutamated at all (e.g., PT523). SPECIFIC AIM 3: To synthesize a series of second generation analogs of 2-desamino-2-methylaminopterin (PT557, dmAMT), which was previously shown to be a weak DHFR inhibitor as the monoglutamate but a potent 'multitargeted antifolate' once it is metabolized intracellularly to polyglutamates. Because only the polyglutamates, and not the parent drug, will potently inhibit the target enzyme these compounds will more truly represent 'prodrugs' than AMT, and should therefore display improved selectivity against tumors with high FPGS activity relative to dose limiting host tissues. The focus will be on enhancing polyglutamation of the lead compound dmAMT by modification of the B-ring and/or pABA moiety.